Package management tools

Poetry, pip, and conda are all package management tools in Python.

• pip is a simple, lightweight tool for installing and managing Python packages.

• conda manages both Python and non-Python dependencies (e.g. c libraries), as well as environments.

• Poetry also manages dependencies and environments, but only for python. It focuses on re-producibility by managing the versions more precisely and provides ease of use

1. pip

• Purpose: pip is the default Python package installer. It is used to install and manage packages from the Python Package Index (PyPI).

• Installation: pip installs Python packages globally or within a virtual environment.

• Package Management: It installs and manages Python packages, and resolves dependencies between packages, but it does not automatically manage Python versions.

• Requirements File: pip uses requirements.txt to list dependencies, but it doesn't automatically lock dependencies to specific versions (unless you manually pin them).

• Environment Management: pip alone doesn't handle environments; it typically works with tools like venv or virtualenv to create isolated environments.

2. conda

• Purpose: conda is both a package manager and an environment manager. It can handle Python packages, but it can also manage packages from other languages (e.g., C, R, Julia).

• Installation: conda installs packages from its own repositories (Anaconda or Miniconda).

• Package Management: Unlike pip, conda can manage not only Python packages but also non-Python dependencies (e.g., C libraries, system-level tools).

• Environment Management: conda can create and manage isolated environments for different projects, which include both Python packages and non-Python dependencies.

• Requirements File: Conda uses environment.yml files, which specify environments, dependencies, and versions.

3. Poetry

• Purpose: Poetry is a dependency manager for Python, aimed at simplifying packaging and distribution, as well as managing project dependencies and environments.

• Installation: Poetry is installed as a standalone tool (not included in Python by default).

• Package Management: Poetry simplifies dependency management by automatically handling virtual environments and resolving dependencies with a focus on locking versions for reproducibility. It uses pyproject.toml (instead of requirements.txt) for defining project dependencies.

• Environment Management: Poetry automatically creates virtual environments for projects. It handles both the installation of dependencies and the creation of environments, removing the need for an external tool like venv.

• Requirements File: Poetry uses pyproject.toml to list dependencies, as well as poetry.lock to lock versions of packages, ensuring reproducibility across different environments.

• Use Case: Poetry is great for managing Python projects, especially when version control, reproducibility, and simplicity are important.